Fairness in a one-principal-two-agents game - a post-experimental questionnaire analysis

Abstract We analyze the data of a questionnaire which students filled out after participating in a principal–agent experiment featuring a single principal and two agents. Each agent’s choice of (costly) effort determines a return that is split between principal and the respective agent according to a payment rule which the parties have agreed upon previously. Guth et al. [Schmalenbach Bus. Rev. 53 (2001) 82] show that the observed distribution of earnings between players is less asymmetric than economic rationality proposes. This can be seen as indirect evidence for fair behavior. The post-experimental questionnaire directly asks what participants consider as fair in that game. The responses reveal that fairness perceptions are based on the group of three players (triad) rather than the dyad between principal and agent. Fairness judgements reflect productive asymmetries between players and, more surprisingly, differences in strategic power. By and large the players agree on the just order of payoffs, despite some egocentric bias. Actual experimental decisions and questionnaire responses are strongly correlated.